Voice recognition framework for open rich-client mobile applications

The presentation layer, in the pattern for mobile client application architectures, is responsible for managing the user interface. For the case of open systems, the components of this layer must be adapted to the capabilities and constraints of mobile devices for processing. The ability of an open application process can be adapted for each mobile client applications generated the concept of 'rich-client'. The rich-client applications always have the presentations layer and business layer, inside the presentation layer are the UI components. Traditionally the presentation layer has the processing and the UI component, however the advantage that a rich-client presents is that the device or an external service can execute processing. The main advantage to execute the process inside the device is to use the features and resources of mobile device. In rich-client applications are intended to API's and frameworks UI components are open and processing is performed, if possible, taking advantage of the device's resources. This paper presents the implementation of the speech recognition component for rich-client applications. The API definition of this component is based on HTML5 and JavaScript. For the processing of this component has been made vocal recognition prototype that runs natively on Android and iOS devices. The processing of this component has been tested in a prototype based on a recognition of vowels that runs natively on Android and iOS devices.